New evidence-based treatments have been slow to be adopted into the routine practice of addiction treatment, and traditional methods for training community-based clinicians in such new treatments are of limited effectiveness. In this renewal, we propose to continue development of methods for training clinicians in Motivational Interviewing (MI), based on the principle of live supervision. Live supervision (supervisor and trainee seeing a patient together) is common in basic training, but rarely used in continuing education. In the initial funding period we developed Teleconference Supervision (TCS), which implements live supervision by an expert through teleconferencing technology, harnessing the principles of context-dependent learning, differential reinforcement with immediacy of feedback, and modeling. Clinicians interview standard patients at their home clinics, while Supervisors listen via telephone. The clinician wears an earpiece, through which the Supervisor provides immediate feedback and coaching designed to differentially reinforce and shape MI skills. In a randomized controlled trial with community clinicians (N= 100), TCS produced superior overall MI skill, compared to control conditions where clinicians received Workshop only training, or standard tape-based supervision. Findings were encouraging, yet the proportion of clinicians achieving expert proficiency was limited, and there were trends toward loss of skill at follow- up, suggesting longer training is needed to maximize and sustain MI skill. In the new funding period, we propose to test an improved version of Teleconference Supervision (TCS) that will preserve and refine the initial 5 live supervision sessions and will add 4 booster sessions over the 2 months after completing the initial TCS training. For the booster sessions, the Clinician will have previously audiotaped a session with an actual patient, and Supervisor and Clinician will be able to listen together to the session over the telephone, while discussing it. While both live TCS and tape-based booster sessions offer context dependent learning, the live sessions provide for modeling and differential reinforcement with immediacy of feedback, and tape review boosters provide ample time to role-play and discuss technique and strategy. The proposed project will first refine TCS in a Stage 1a pilot trial, then move to a randomized Stage Ib trial, with Standard Supervision as the control condition. We hypothesize that TCS will produce superior MI skill among community-based Clinicians, and more change talk and improved substance use outcome among patients. If it continues to show promise, methods like TCS could re-shape the field in terms of disseminating new treatment techniques. Future directions could include larger Stage 2 or 3 trials of TCS, adaptation of TCS to other treatment techniques, and development of group supervision models to increase sustainability and make more efficient use of expert supervisors' time.